feat(e2e): render direct/recorder/debug/tcplogger/stress sections in the unified report

Extends report_build.py with build_by_kind() (per-scenario-kind pass/fail rollup) and a ported build_stress()/stress_headline()/stress_plots() (scaling curves per stress axis), wires both into the headline KPIs, regression tracking, and report_data.json. E2E_Report.typ renders the four new per-kind tables plus a Stress Tests section (per-axis case tables + scaling plots, gracefully degrading to placeholders when a kind/stress data is absent). run_e2e.sh now passes --stress-results so the report actually receives real stress data instead of silently omitting it. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-07-01 21:57:11 +02:00
parent b65ac06ce2
commit 8337d678be
4 changed files with 235 additions and 5 deletions
@@ -323,6 +323,79 @@ MARTe2 processes, plus sustained client throughput (recorded samples ÷ duration
  #v(6pt)
 ]

+// ── per-kind sections (direct/recorder/debug/tcplogger) ─────────────────────
+// Raw scenario records here come straight from results.json (not e2e.scenarios'
+// reshaping), so they only carry id/kind/status/known_issue/metrics — no
+// waveform-fidelity breakdown (already covered, where applicable, in the
+// Scenarios section above for chain-kind scenarios; these kinds run through
+// dedicated non-chain harnesses in run_e2e.sh).
+#let kind_table(title, block) = {
+  [= #title]
+  [#block.n_pass/#block.n_total passed.]
+  v(4pt)
+  if block.n_total == 0 {
+    text(fill: neutral)[_No scenarios of this kind in this run._]
+  } else {
+    table(
+      columns: (1.4fr, 0.8fr, 2fr),
+      align: (left, center, left),
+      stroke: 0.4pt + rgb("#d0d7de"),
+      inset: 5pt,
+      table.header([*Scenario*], [*Status*], [*Known issue*]),
+      ..block.scenarios.map(s => (
+        [#s.id],
+        status_badge(s.status),
+        if s.at("known_issue", default: none) != none {
+          text(size: 8pt, fill: rgb("#7d4e00"))[#s.known_issue]
+        } else { text(fill: neutral)[—] },
+      )).flatten()
+    )
+  }
+}
+
+#kind_table("Direct Round-Trip (UDPStreamer ↔ UDPStreamerClient)", data.direct)
+#kind_table("Recorder (BinaryRecorder disk output)", data.recorder)
+#kind_table("Debug Service E2E", data.debug)
+#kind_table("TCPLogger E2E", data.tcplogger)
+
+// ── stress tests ─────────────────────────────────────────────────────────────
+= Stress Tests
+#if data.stress == none [
+  _Not run this session._
+] else [
+  #align(center, status_badge(data.stress.overall) + h(8pt) + text(size: 11pt)[
+    #hl.at("stress_pass", default: 0) passed ·
+    #hl.at("stress_fail", default: 0) failed
+  ])
+  #v(4pt)
+  #for (axis, cases) in data.stress.by_axis [
+    == Axis: #raw(axis)
+    #table(
+      columns: (0.8fr, 0.8fr, 1fr, 1fr, 1fr, 1fr),
+      align: (right, center, right, right, right, right),
+      stroke: 0.4pt + rgb("#d0d7de"),
+      inset: 4pt,
+      table.header([*Level*], [*Status*], [*Hub RSS (MB)*], [*MARTe RSS (MB)*],
+                   [*Zoom p50 (ms)*], [*Zoom p95 (ms)*]),
+      ..cases.map(c => (
+        [#c.at("level", default: "n/a")],
+        status_badge(c.at("status", default: "?")),
+        fnum(c.at("hub_rss_mb", default: none), digits: 1),
+        fnum(c.at("marte_rss_mb", default: none), digits: 1),
+        fnum(c.at("zoom_p50_ms", default: none), digits: 1),
+        fnum(c.at("zoom_p95_ms", default: none), digits: 1),
+      )).flatten()
+    )
+  ]
+  #if data.stress_plots.len() > 0 [
+    #v(4pt)
+    == Scaling curves
+    #grid(columns: 2, gutter: 8pt,
+      ..data.stress_plots.map(p => image(p, width: 100%))
+    )
+  ]
+]
+
 // ── trend plots ──────────────────────────────────────────────────────────────
 #if data.trend_plots.len() > 0 [
  = Trends over runs